The Divvy Density Dilemma: Are Stations in Low-Income Areas Too Far Apart?

This station by Kennedy-King College in Englewood is a 3/4-mile walk from neighboring stations. Photo: John Greenfield

Planning a useful, equitable, and financially sustainable bike-sharing system in a big, diverse city like Chicago is no easy task. You have a finite budget, and therefore a limited number of cycles and docking stations to work with. You want to provide access to the system for as many people as possible, and you’re certain to get complaints from residents and politicians whose neighborhoods don’t get bikes. However, if you spread the available stations across too large a service area, there will be poor station density and the system won’t be convenient to use.

I respect the the fact that the Chicago Department of Transportation has had to make some tough decisions in implementing the Divvy bike-share system. However, a new study from the National Association of City Transportation Officials suggests that the city may have made a mistake by placing Divvy stations too far apart from each other in many neighborhoods, especially low-income communities. The report, titled “Walkable Station Spacing Is Key to Successful, Equitable Bike Share,” argues that cities don’t do residents any favors by creating sprawling service areas that cover large numbers of neighborhoods, but don’t provide a useful network.

Low station density discourages use and undermines equity

The NACTO paper notes that, while bike-share can be an inexpensive, time-saving form of transportation, low-income people are underrepresented among American bike-share customers. In the U.S., poor neighborhoods tend to have a relatively low density of people and destinations, and when bike-share planners respond to this by putting a lower density of stations in these communities, it exacerbates the usage issue.

The study argues that, just as people usually aren’t willing to walk more than ten minutes to a rapid transit stop, if bike-share stations are located more than a five minute walk from a person’s starting point or destination, that person will generally choose a different mode. That jibes with my personal experience. I’m fortunate to live a quarter mile away from a Divvy station, but I find the five-minute walk to and from the station a little annoying, and if it was another block away I’d probably use it less often.

NACTO’s analysis of several different North American systems supports the five-minute rule theory. They found that the number of rides per day to or from a given station increases according to its proximity to other stations. For example, bikes in New York’s Citi Bike system, with 23 stations per square mile, got more than three times as much use as those in as the Twin Cities’ Nice Ride network, with only four stations per square mile.

Therefore, NACTO recommends that stations be placed no more than a five-minute walk from each other, which they define as 1,000 feet, for a density of 28 stations per square mile. I’d argue that average walking speed is a 20-minute mile, so placing stations every quarter-mile (two standard Chicago blocks), for a density of 25 per square mile, should be sufficient.

Low-income people tend to have less spare time and disposable income than wealthier folks, so they are even more likely to be deterred from paying to use bike-share if the station locations aren’t convenient. The study argues that, while efforts to increase bike-share use by low-income people have focused on offering discounted memberships and providing access to unbanked individuals, the density issue has largely been overlooked.

NACTO recommends having a consistently high station density across the service area, including poor neighborhoods with relatively low population densities. Rather than reducing the number of stations in these communities, the number of docking points at the stations should be adjusted according to demand.

How the city approached the 2013 Divvy Rollout

Old Divvy
The 2013 coverage area is shown in Blue, and the new service area is in red. Most of the areas with quarter-mile spacing are densely populated and relatively affluent, while low-income communities generally received half-mile spacing.

Let’s look at how the Chicago Department of Transportation has approached equity and station density. When they installed the first batch of 300 stations in 2013, the service area was split fairly evenly between the South Side, which contains many low-income communities, and the North Side, which is generally more affluent. Few neighborhoods on the generally low-income West Side got bikes.

The service area was 44.1 square miles, with an overall station density of 6.8 stations per square mile, including industrial zones and other non-populated areas that didn’t get stations. Station density is somewhat higher when you exclude those areas – the NACTO report, written before the current expansion, lists Chicago’s station density as 8 per square mile.

In this first round of installations, CDOT chose to concentrate the stations downtown and on the North Side, where stations were placed roughly every quarter-mile. In these areas, you’re usually within a five-minute walk from a station, so the system is convenient to use.

However, south of Roosevelt Road, stations were generally spaced every half mile, or four blocks. This southern portion of the original service area includes low-income communities like Bronzeville, Douglas, and Oakland. When you have half-mile spacing, 50 percent of the territory bounded by any four stations is located more than a five-minute walk from a station. If you’re unlucky enough to live in the center of the quadrant, you’ve got a deal-breaking ten-minute walk to a station. This has been a factor in why Divvy usage has generally been low in these communities.

Predictably, there were complaints about CDOT’s strategy. Some people who got only a sprinkling of stations in their communities felt short-changed. And others who didn’t get Divvy in their neighborhoods at all were even more upset.

CDOT officials said they needed to provide a higher density of stations downtown and on the North Side because those areas have a higher density of residents and destinations like ‘L’ stations, job centers, and retail businesses. They argued that, if they didn’t take advantage of the higher potential for use in these areas, the system wouldn’t be financially sustainable, and they’d never get the chance to bring bike-share to more neighborhoods.

Chicago’s current bike-share expansion

About two years later, the city is now making good on its promise to provide service to more communities, adding 176 more stations and nearly doubling the service area to 86.7 square miles. The number of wards served is growing from 19 to 33, and many low-income neighborhoods are gaining access to the system, including Woodlawn, Washington Park, Englewood, Canaryville, Little Village, Lawndale, and East Garfield Park. And instead of CDOT concentrating stations in the denser or wealthier parts of the new service area, the stations are being equally distributed.

Sounds like a recipe for equity, right? The problem is that, because the service area is expanding to such a broad area, with a relatively small number of new stations, the overall station density is dropping to 5.49 stations per square mile, including unpopulated areas. Almost all of the new areas, irregardless of population density or income level, are getting substandard, half-mile station spacing. Meanwhile, downtown and Hyde Park – just about the only neighborhood with a high bike mode share that got a low station density in the previous round – have received additional infill stations.

The upshot of CDOT’s strategy is that Chicago will remain a city with two different kinds of bike-share neighborhoods. The system will continue to flourish downtown and in the relatively affluent North Side neighborhoods that received quarter-mile spacing in 2013. However, usage will likely lag behind in most of communities that are getting half-mile spacing, especially in the low-income areas that had a low bike mode share to begin with.

6th Ward Alderman Roderick Sawyer has already expressed concerns about just how useful the six, widely spaced stations in the low-income Englewood community will be. “I mean, where are you going that has another Divvy station?” he told DNAinfo. “To that extent, I just want to make sure they get good use, that’s all. If they’re going to be anywhere, they need to be everywhere.”

CDOT doesn’t necessarily deserve blame for this outcome. As stated earlier, creating a bike-share system is a balancing act. If they had done half-mile spacing downtown and on the North Side, the entire system would have seen much less use, and might have been a fiscal flop. On the other hand, the agency has received a lot of political pressure to quickly expand the system to new neighborhoods. Perhaps they felt they had no choice but to resort to low-density placement in order to reach these areas.

How the city could have handled station placement and density

Hypothetical Divvy service area if all 476 stations were spaced 1/4 mile apart
Here’s what Divvy coverage might look with 476 stations and quarter-mile spacing. Map: Steven Vance and John Greenfield

While it’s way too late to completely reconfigure the Divvy system, let’s consider what Chicago’s coverage might have looked if the city had used quarter-mile spacing across the entire service area and, with an eye on geographic equity, made the system radiate out evenly from the Loop. As you can see from the adjacent map, which takes into account unpopulated areas, the 476 stations would occupy a semicircle with a five-mile radius.

This strategy would have provided coverage to a little over 39 square miles of the central city, a much smaller area than the planned 86.7 square-mile service area. However there would have been consistently good access across the parts of the city that were covered. As prescribed by NACTO, neighborhoods with low population density, including low-income areas, would get fewer docking points per station, but the stations would still be a convenient walking distance from each other.

The road ahead

If it’s the case that CDOT made a mistake by not providing proper station density in most of the service area, what should they do next? Mayor Emanuel has pledged to expand Divvy to 6,000 bikes, i.e. 600 stations, within the next four years. Aldermen in outlying wards have already said they are frustrated that their constituents are getting passed over in the current round. Therefore, the path of least resistance would be to simply repeat the error of expanding the system as quickly as possible, continuing with the substandard half-mile station density.

But maybe a better choice would be to slow down the expansion and use most of the next 124 stations as infill, increasing density in the underserved parts of the existing service area. The focus could be on low-income neighborhoods where residents stand to gain the most from the economic and health benefits of cycling. That might be the best course of action, if we really care about making the system more equitable.

  • Kevin M

    Nice article.

    Yes, John’s right: infill, and ignore the cries from the outliers.

    To the outliers: There’s a much larger and older inequity in the CTA train line coverage that needs your attention. Your tax dollars support the CTA (to a far greater expense than Divvy), so you should should be shouting for bus rapid transit (e.g. Ashland BRT), extension of existing lines, and new train lines to bring your neighborhoods access to fast, reliable transit like others have long had access to in our city. Let the bike-share system get up and running properly before demanding blanket coverage to all corners of the city.

  • dd

    Not sure. The NACTO study seems pretty flawed, as they have no methodology for teasing out causes and are presenting some highly correlation-driven results. It would be relatively easily to control for population density and then determine the effect of station density, but they haven’t taken that simple and pretty essential step.

    I would also like like to see the stations controlled for commercial density. < a pretty great analysis, I thought.

    Control for population density, and translate "bike trips per bike" into "bike miles per bike," and then there is a discussion to be made. As is, the conclusions don't seem ready for prime-time policy-driving.

    Expanding to low-income areas strikes me as more worthwhile than optimizing one form of usage, but having that discussion coherently requires a better understanding of what underlying factors actually drive use, and a smart, shared definition of what "use" is. I don't think NACTO gets us there.

  • Anne A

    To give one specific example, on 43rd St. there is a Divvy station next to the green line, where there are few adjacent businesses. The next station to the east is at Cottage Grove, almost 1 mile away. There are businesses in the middle of that stretch (especially Ain’t She Sweet Café and Sip and Savor Coffeehouse) that could benefit from a Divvy location, but the walk to/from the nearest Divvy station might be daunting for many potential customers at those cafes.

  • MLKendricks

    I’m still trying to parse out my thoughts on this:

    I think one thing that didn’t get enough attention in NACTO article is the financial points. $75 is not a lot of money for most people, but for people who are relatively new to biking, with not a lot of cash in the first place $75 might be too big of a leap of faith to try on bike share. The Credit/Debit Card requirement I also don’t think gets enough credit. If the areas are low income, they are more likely to be unbanked, further depressing potential ridership >>>
    If Divvy is densifying in low income areas, I don’t think they can just add more stations and expect more usage, if the other reasons (especially financial reasons) aren’t addressed. I don’t know what discounts Divvy is offering for low income or cash payment options (Like Arlington >>> ), but that outreach should be concurrent with expanding more bike share stations into low income areas.

  • From a SBC post a couple weeks ago based on an interview with CDOT’s Sean Wiedel:

    Despite these outreach efforts, the bike-share system’s requirement that users have a credit card will continue to be a barrier for unbanked Chicagoans and limit use in low-income areas. Wiedel promised that CDOT is taking concrete steps to address this issue, and will have a major announcement by early summer.

  • JacobEPeters

    Alternately, the next year will provide useful data as to what stations are being used & how frequently in each neighborhood. If CDOT stated that stations had to stay in their community area or ward, but that they could move depending on frequency of use, we could see stations relocated to higher densities in parts of the neighborhood that embrace the system, building off of successes. There are a lot of downsides to this as a strategy, but if it was clear from the start that stations would be judged in this manner, but that they would not be moved from one side of town to the other, then maybe it would have a chance.

  • Biker

    Completely agree, and one of the reasons why I didn’t renew my membership. Stations are too far apart for me (I live in lakeview), and docks were often full. To me, it makes no sense to place stations in the fringe areas, 1/2 mile apart. Where does CDOT think people will ride those bikes to?

  • So you’re saying that the quarter-mile spacing in Lakeview (which means you’re never more than two standard Chicago blocks, a five-minute walk, from a station) wasn’t dense enough for you?

  • Jared Walker of Human Transit addresses the core philosophical issues regularly. It’s one of the core decisions a community must make about what kind of transit we want. This core issue here is the coverage versus ridership one. The community must decide whether to emphasize ridership or coverage. ( ) Now with bike share the issues overlap a bit. Stations that are too far apart compromise both the coverage and ridership goals at the same time.

    A lot depends upon the overall success of the system. If enough tourists can subsidize enough bike/station density in enough good locations then enough usage there can subsidize the expansion into less optimum locations. etc.

    Will there be tokenism. For sure but putting too many stations too densely in places where there is not the ability or the desire to use them would be counter-productive for everyone.

  • This by-the-way is the same core issue as the one the Lincoln bus (and others). Do we want coverage as priority or do we want ridership as a priority.

  • Cecilia Gamba

    I think the expand-then-infill strategy is a pretty good one to balance those issues. Yes, if there are just a few stations in your neighborhood that are not very conveniently located for your travel patterns you probably won’t get a membership. But, the stations will be convenient for *someone* — maybe more for last-mile connection to transit than for moving around the neighborhood. So you’ll start seeing some bikes around, you’ll get familiar with the system.. Then comes infill (“subsidized” by the higher-use areas) and you’re more likely to be ready to join. It will be interesting to see the results of the Hyde Park infill from
    this round (although it’s not a low-income community). Other parts of the South Side did not get infill, and in the future you certainly need more than 124 stations to infill AND expand further, but I think the general strategy is better than an all-or-nothing approach, which would end up either less equal (if you focus on high-use areas), or less financially viable (if you maintain geographic equity).

  • Erik Swedlund

    That IS a really great analysis at TheWashCycle, thanks for pointing it out.

  • ohsweetnothing

    I can’t be the only one that sees the cognitive dissonance of the same elected officials that are demanding Divvy expansion to the far outlying neighborhoods being the same elected officials critical of policies that would actually make these same neighborhoods more friendly to those on bikes. Speaking out against RLCs and Speed Cameras, attempting to raise the speed limit on City Streets, attempts to lower the price of City Stickers, general criticism of spending City dollars on bike infrastructure (including Divvy initially!), etc…

    If the complaints are a sign of a shift in attitude towards the biking as a viable means of transportation in Chicago, then I can get behind this. If it’s simply wanting a resource just because it exists, then the complaints all ring hollow to me. You’ve got to be all in or all out re: bikes in Chicago.

  • Fedor Manin

    It’s not mere density that’s important; the stations have to be in the right places, as in where people want to go and not the next block over. Divvy is most useful for very short, 1-2 mile trips, and having to walk even a block makes it far less so. I would like to use Divvy to get from the Red Line to Hyde Park, and the fact that stations have been installed a block away in the wrong direction from the Red Line stations is a huge disincentive.

    Or say I want to get from the Fullerton Red Line stop to the Broadway and Diversey intersection in Lakeview. If there were stops actually at either end, I would use Divvy; the fact that the stops are half a block and two blocks away, respectively, tilts the scale towards walking.

    Just measuring density makes sense in residential areas where there aren’t many popular destinations. I think people are a little more willing to walk from their houses to a station.

    I think Divvy has tended to place stations where convenient, ignoring the need to have as many destinations as possible within oh say a 200-foot radius in order to maximize the usefulness of the system.

  • I agree that it would be great if Divvy stations could be placed in more convenient, intuitive locations. Ideally, when you walk out of an ‘L’ station, the Divvy station should be right there, rather than around a corner on a side street half a block away, so that you have to pull out your smart phone to find it (I’m looking at you, North/Damen Blue Line stop Divvy station.)

    And it’s always baffling when a station is installed at an intersection where two or more of the corners are occupied by a gas station, a parking lot, a vacant lot, or a cemetery. But there are probably a lot of factors in CDOT’s placement decisions that aren’t obvious to us, such as pedestrian movements, the availability of sunlight for the solar panels, the whims of the local alderman, etc.

  • If the system can’t garner citywide political support it won’t survive and prosper. Accordingly, I don’t get too worked up about the compromises that needed to be made to get Divvy stations in more wards. We’ll get to the right place in the end on density as more in-fill stations are placed.

  • Lakeview Guy

    No, it wasn’t. I’ll give you an example – one night I grabbed a bike at the Belmont CTA, hoping to park it at Broadway and Belmont. B&B was full, as was Halsted and Cornelia, and Halsted and Roscoe. I ended up taking the bike back to the Belmont CTA and walking home. Do you think I’m going to renew my membership and see Divvy as reliable and useful? I don’t think so.

  • OK, sounds like the problem in that case wasn’t station density but a need for more docking points at those locations and/or better rebalancing. Fortunately so far in most of Chicago, that sort of thing seems to be an exception to the rule — I haven’t heard many complaints about it or experienced it much myself.

  • I wish there were a big red “I tried to dock here and couldn’t” button on the kiosk one could push in that situation, so Divvy can get failed-attempt data and know they need to build out capacity.

  • Lakeview Guy

    True – I guess my line of thinking was if there were more stations,there would likely be more docks. Great idea about a big red “i tried to dock here” button. Would provide great data to show where stations could be expanded or better balanced.
    But unfortunately for me, once bitten, twice shy. I’ll walk/drive/CTA/ride my own bike and won’t support Divvy any more.

  • what_eva

    Agreed. I know there’s some way to get an extension to your time. Never used it, but I presume it’s on the touchscreen. But if you don’t need the time extension, you’re not going to stop and mess with the screen. A big button to push would solve that.

  • what_eva

    I get sick of hearing that this is all a white/high income vs black/low income issue. The lakefront neighborhoods are very, very dense, that’s why they get the coverage they do. You don’t see 1/4 mile coverage in Old Irving Park or Edgebrook or Beverly or numerous other white/high income neighborhoods that lack density.

  • Far-flung Edgebrook and Beverly (which is 34% African-American) probably won’t be getting stations for years. It will be interesting to see what happens in the relatively affluent neighborhoods with higher bike mode share that are getting low station station density in this round of installations. If the Hyde Park pattern is repeated, they’ll get infill stations in the next round.

  • what_eva

    I just think it’s CDOT being practical and putting more stations where there is population density. I feel like you’re ignoring that (I *know* the whining alderman are ignoring that). It’s less about income or race and more about population density.

  • The NACTO study argues that, regardless of population density, you need to provide consistent station density across the service area, or else the system is doomed to fail in the low-density, often low-income, neighborhoods. They say the way to address varying population density is by adjusting the number of docks at each station.

  • what_eva

    The question becomes, at what population density does bike share simply not work? Clearly there’s a bottom somewhere (ie you wouldn’t put 1/4 mile station density in farmland). Are some of the outlying neighborhoods below that minimum population density?

    I think the better answer is probably your suggested map.

  • trufe

    thanks for the advice about what types of transit we can rightfully seek. pretty sure most of us are aware of the unfair train service allocation,

    of course, the rationale behind those long time low service levels for other types of transit are the exact same as the one you now posit. “let it get up and running properly downtown/in higher density areas first, and we will get to you eventually.”

    the problem is they never do “get to” us. how long has the red line extension been in the long term plan? even if the initial phase of the ashland BRT eventually gets built and is successful, do you really think it will ever be extended down to 95th?

    no, we will be told they are going to expand BRT to halsted or western, or maybe some east west streets like chicago or jackson to “really get the system up and running” before they extend it further south (or probably much farther north for that matter)

    not to mention, of course, that it is a self fulfilling prophecy because higher density is not going to happen without better, useful transit options

    if ANYTHING is ever going to make its way out, it would be the relatively inexpensive DIVVY, not a pie-in-the-sky train line, or even BRT

    so i think i will choose to ignore your advice, in my typical, naive, uninformed, outlying area way.

  • This is the big issue here. Articles like this continue to get churned out of the mill beating up on bike share systems for not being equitable while ignoring the bigger elephant in room: horrific street design. Even this very blog has published several pieces critical of the unsafe infrastructure in low-income areas. At the same time, an enduring argument continues to be that low-income folks seem to be the ones who would benefit the most from bike share systems because they already bike the most too.

    The disconnect there is mind-boggling: if disadvantaged individuals currently bike the most, they probably need infrastructure to bike on, not fancy bikes to ride on the same crappy streets. It would in general almost certainly be better to use the money to fix the streets first, including the addition of bike parking at bus stops. That’s a far cheaper approach to the first-/last-mile problem than Divvy or any other bike share system. Only after that process is complete can chasing money after bike share really show a better value to the residents of these areas.

  • Jared Kachelmeyer

    Some sidewalks are probably simply too narrow. They do put them in road space but I doubt that’s always an option especially if the space in front of a station is a bus stop or a traffic lane.


Why Divvy Needs to Densify as It Expands

For a very new American bike-share system, Divvy is doing well, but it has a lot of room to improve, according to the Institute for Transportation and Development Policy’s new Bike Share Planning Guide [PDF]. The guide includes best practices for designing, distributing, and marketing this new form of transit. While Divvy is still growing […]

The Pros and Cons of Divvy’s New Expansion Map

In mid-July, Alta Bicycle Share’s Mia Birk told Marketplace that, due to pipeline issues, new bikes for the systems Alta runs probably wouldn’t arrive until 2015. At the time, I predicted that the Chicago Department of Transportation would soon announce that it wouldn’t be able to expand the Divvy system from 3,000 to 4,750 cycles […]

Diving Into Divvy Stats: Bike-Share Trips Spike on the Weekend

Steven Vance is the self-described “data geek” at Streetsblog Chicago, but even a right-brained type like myself couldn’t help but be intrigued by some recently revealed stats and charts about Divvy bike-share use patterns. Chicago Department of Transportation Deputy Commissioner Scott Kubly shared them during a talk at last week’s Complete Streets Symposium, hosted by […]

Correcting Konkol: South and West Sides Received the Bulk of New Bike Lanes

Just because you’ve won a Pulitzer Prize doesn’t mean you always get your story straight. DNAinfo’s Mark Konkol posted a column today arguing that Mayor Rahm Emanuel’s bike initiatives are a frivolous distraction from the city’s pressing crime, education, and budget issues. He also claims that Emanuel’s administration hasn’t been giving low-income communities their fair share of […]